Specific Aim 3: To evaluate the impact of self-reported sedentary behavior on GWG and meeting recommended guidelines for GWG among high-risk prenatal care patients.
Univariate Analysis: Descriptive analyses will be conducted as under specific aim 1. Numbers and frequencies will be calculated for categorical variables and means and standard deviations for continuous variables (Tables 1-3).
Bivariate Analysis: Bivariate analyses will be used to cross-tabulate baseline characteristics by lower versus top three quartiles of baseline and post-intervention sedentary behavior and change in sedentary behavior from pre- to post-intervention (Table 21+22) as well as by outcome group as described for aim 1 (Tables 4-7) using Student’s t test for continuous variables and Pearson’s chi square or Fisher’s exact test for categorical variables.
To account for the multi-level outcome of meeting IOM guidelines (inadequate, adequate and excessive GWG), multinomial logistic regression will be used to conduct an initial unadjusted analysis of the association between baseline and post-intervention sedentary behavior, as well as change in sedentary behavior, and odds of meeting IOM guidelines for total weight gain during pregnancy and meeting IOM guidelines for rate of weight gain in 2nd and 3rd trimesters. Linear regression will be used to conduct an initial unadjusted analysis of the association between baseline sedentary behavior,
post-intervention sedentary behavior and change in sedentary behavior and total weight gain during pregnancy and rate of weight gain in 2nd and 3rd trimesters as continuous variables (Table 23).
Multivariate Analysis: Multivariable adjusted analyses will be conducted to assess the impact of baseline and post-intervention sedentary behavior and change in sedentary behavior over the course of the intervention on meeting IOM guidelines, as well as total weight gain during pregnancy, and rate of weight gain. This analysis will be adjusted for potential confounding factors, and all factors resulting in a change in more than 10% of the main effect estimate will be included in the final model (Table 24). Models
including post-intervention or change in sedentary behavior will by default adjust for baseline sedentary behavior.
Age (<30 years, ≥30 years), ethnicity (Hispanic, non-Hispanic), pre-pregnancy BMI (<30 kg/m2, ≥30 kg/m2) and parity will be evaluated as effect modifiers by including interaction terms in multiple logistic and multiple linear regression models.
For any potential effect modifiers that result in a significant likelihood ratio test for the full versus reduced model, the results will be stratified by levels of the effect modifier.
Power and Sample Size
A total of 290 women were initially randomized into the B.A.B.Y. Study. Of those who enrolled, based on data from the 2010 survey by the Pregnancy Nutritional Surveillance System, we expect 48% to have excessive GWG and 21.5% to have
inadequate GWG according to IOM guidelines (85). Given an expected loss to follow-up of 15%, the average loss to follow-up in a similar randomized controlled trial conducted
by Luoto et al. (41), we expect a final study population of 246, with 118 women exceeding, 75 meeting and 53 not achieving IOM guidelines. In this study population, with a ratio of 1:1 in the exercise group as compared to the health and wellness group, the power to detect a range of odds ratios for exceeding versus meeting IOM guidelines, as well as for not achieving versus meeting IOM guidelines are presented in table 25, assuming a 0.05 two-sided significance level.
In a randomized control trial of a dietary intervention conducted by Asbee et al. in a similar population in the United States, women in the control group had an average GWG of 35.6±15.5 lbs (131). Given the above 1:1 ratio of exposed to unexposed, the power to detect a range of mean differences in GWG is presented in table 26. This study has more than 80% power to detect a reduction in total GWG of 6 pounds, which, given the estimate of average GWG of 35.6 lbs from Asbee et al, would result in an average GWG of 29.6 in the intervention group.
Based on an expected 1:2 ratio of participants who increase/maintain verses decrease their moderate/vigorous and sports/exercise activity, the power to detect a range of odds of exceeding IOM guidelines versus meeting IOM guidelines and not achieving IOM guidelines versus meeting IOM guidelines for those increasing/maintaining their activity versus those decreasing their activity is presented in table 27. The power to detect a range of mean differences in GWG for those who increase/maintain versus decrease their activity is presented in table 28. Similarly, assuming a 1:2 ratio of
participants who decrease/maintain versus increase their sedentary behavior, the power to detect a range of odds of exceeding IOM guidelines versus meeting IOM guidelines and not achieving IOM guidelines versus meeting IOM guidelines is presented in table 27 and
the power to detect a range of mean differences in GWG for those who decrease/maintain versus increase their activity is presented in table 28.
Results
Study Population Characteristics
Of the 488 participants initially enrolled in the study, 290 met study inclusion criteria and were randomized into the exercise or health and wellness groups. After randomization, participants were additionally excluded from the final analysis if they had medical contraindications to exercise, experienced a miscarriage/termination of
pregnancy or were missing data on delivery weight, leaving a final study group for analysis of 241 participants (Figure 1).
A total of 49.4% of participants were randomized into the exercise group and 50.6% into the control group. At baseline, median sports/exercise activity was 4.4 (IQR 0.4, 11.9) MET-hrs/wk and median moderate-vigorous activity was 65.7 (IQR 28.2, 128.5) MET-hrs/wk (Table 1, Figure 2). Sports/exercise increased from baseline to post-intervention in the exercise group (median= 4.8; IQR: 0.0, 13.5 MET-hrs/wk) and remained unchanged for moderate-vigorous activity (median = 0; IQR: -4.3, 3.6 METhrs/wk). Moderate/vigorous activity decreased slightly in the exercise group (median = -1.2; IQR -51.2, 32.4 MET-hrs/wk) and decreased in the health and wellness group (median = -18.6; IQR: -66.2, 5.3) (Table 1, Figure 2)
Average total GWG was 30.5 (SD 20.0) lbs. Only 16.2% of participants met the IOM guidelines for adequate total GWG; 16.6% had inadequate weight gain and 67.2%
had excessive weight gain. The average rate of GWG during the second and third trimester was 0.95 (SD 0.62) lb/wk. Only 10.4% of participants met IOM guidelines for
adequate rate of GWG, while 19.8% were classified as inadequate and 69.8% as excessive (Table 2).
The majority of study participants were young (49.4% less than 25 years), Hispanic (61.0 %), low-income (54.4% with household incomes less than $30,000/year) and with low to moderate levels of educational attainment (54.3% did not receive
education beyond high school). Most participants reported living with a partner (58.9%) and one-quarter were married (25.3%). The majority of participants were parous (72.2%) and overweight (35.3%) or obese (61.8%). One-tenth had a history of GDM in a prior pregnancy (9.5%), and the majority had a family history of type 2 diabetes in a first-degree relative (93.0%) (Table 3).